Model comparison
DeepSeek V4 Pro vs Ministral 3B
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 26.2 on the Noometry Index. Ministral 3B costs 9.9× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.
Last verified . 5 shared benchmarks.
DeepSeek
Mistral AISummary
- They share 5 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 3 categories and Ministral 3B in 0 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek V4 Pro leads 59.5 to 10.4.
- The biggest single-benchmark swing is GPQA Diamond: 91.7% for DeepSeek V4 Pro and 25.3% for Ministral 3B.
- Ministral 3B is cheaper at $0.10 / $0.10 per million input/output tokens, against $0.66 / $1.98 for DeepSeek V4 Pro.
- DeepSeek V4 Pro accepts more context: 1M tokens versus 131K.
Side by side
| DeepSeek V4 Pro | Ministral 3B | |
|---|---|---|
| Provider | DeepSeek | Mistral AI |
| Noometry Index | 54.3 | 26.2 |
| Released | 2026-04-24 | 2024-10-01 |
| Weights | Open | Open |
| Context window | 1M | 131K |
| Max output | 393K | 262K |
| Input $ / M tokens | $0.66 | $0.10 |
| Output $ / M tokens | $1.98 | $0.10 |
| Results tracked | 48 | 6 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
DeepSeek V4 Pro: 52.4 (#34), Ministral 3B: —
| Benchmark | DeepSeek V4 Pro | Ministral 3B |
|---|---|---|
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| LMArena WebDev | 1582 | — |
| SciCode | 51% | — |
| WeirdML | 66.2% | — |
| LMArena Coding | 1470 | — |
| ALE-Bench | 1,403 | — |
Agentic & Tool Use Not comparable
DeepSeek V4 Pro: 32.8 (#58), Ministral 3B: —
| Benchmark | DeepSeek V4 Pro | Ministral 3B |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Vending-Bench 2 | 3,285 | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), Ministral 3B: 18.4 (#282)
| Benchmark | DeepSeek V4 Pro | Ministral 3B |
|---|---|---|
| DTBench | 93.9% | 51.7% |
| LMCA | 45.5% | 5.5% |
| Epoch Capabilities Index | 155.31 | 118.1 |
| ARC-AGI-2 | 61.3% | — |
| Kagi LLM Benchmark | 53.5% | — |
| NYT Connections (extended) | 91.3% | — |
| ARC-AGI-1 | 90.5% | — |
| CritPt | 18% | — |
| Chess Puzzles | 47% | — |
| LMArena Hard Prompts | 1461 | — |
| Mystery Game Puzzles | 43% | — |
| Surface Evolver Bench | 40% | — |
| ForecastBench | 56.1 | — |
Math DeepSeek V4 Pro leads
DeepSeek V4 Pro: 64.8 (#30), Ministral 3B: 26.6 (#258)
| Benchmark | DeepSeek V4 Pro | Ministral 3B |
|---|---|---|
| FrontierMath (Tiers 1-3) | 64.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 76.6% | — |
| OTIS Mock AIME 2024-2025 | 98.6% | — |
| ProofBench | 50% | — |
| LMArena Math | 1455 | — |
| MATH Level 5 | — | 14.4% |
Knowledge DeepSeek V4 Pro leads
DeepSeek V4 Pro: 59.5 (#31), Ministral 3B: 10.4 (#302)
| Benchmark | DeepSeek V4 Pro | Ministral 3B |
|---|---|---|
| GPQA Diamond | 91.7% | 25.3% |
| Vectara Hallucination Rate | 8.6% | 7.3% |
| SimpleQA Verified | 52.9% | — |
| LMArena Expert | 1464 | — |
Multilingual Not comparable
DeepSeek V4 Pro: 54.4 (#45), Ministral 3B: —
| Benchmark | DeepSeek V4 Pro | Ministral 3B |
|---|---|---|
| LMArena Non-English | 1439 | — |
| LMArena Chinese | 1486 | — |
| LMArena French | 1472 | — |
| LMArena German | 1458 | — |
| LMArena Japanese | 1445 | — |
| LMArena Korean | 1447 | — |
| LMArena Russian | 1453 | — |
| LMArena Spanish | 1458 | — |
Instruction Following Not comparable
DeepSeek V4 Pro: 76.1 (#47), Ministral 3B: —
| Benchmark | DeepSeek V4 Pro | Ministral 3B |
|---|---|---|
| LMArena Instruction Following | 1448 | — |
Long Context Not comparable
DeepSeek V4 Pro: 45.0 (#51), Ministral 3B: —
| Benchmark | DeepSeek V4 Pro | Ministral 3B |
|---|---|---|
| CL-bench Life | 13.5% | — |
| LMArena Longer Query | 1458 | — |
Writing & Preference Not comparable
DeepSeek V4 Pro: 65.5 (#46), Ministral 3B: —
| Benchmark | DeepSeek V4 Pro | Ministral 3B |
|---|---|---|
| LMArena Text | 1451 | — |
| LMArena Creative Writing | 1446 | — |
| EQ-Bench Creative Writing | 1553 | — |
| EQ-Bench 4 | 1166 | — |
| LMArena Multi-Turn | 1467 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than Ministral 3B?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 26.2 on the Noometry Index. Ministral 3B costs 9.9× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.
Which is cheaper, DeepSeek V4 Pro or Ministral 3B?
Ministral 3B is cheaper. It lists at $0.10 per million input tokens and $0.10 per million output tokens; DeepSeek V4 Pro lists at $0.66 and $1.98.
Which has the bigger context window?
DeepSeek V4 Pro does, with 1M tokens against 131K.
How many benchmarks do DeepSeek V4 Pro and Ministral 3B share?
5 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Ministral 3B has 6.